My name is Modus Tollens, and I help you spot cheaters.

April 10, 2011 9:40 AM

Subjects don't need formal logic training. They don't need math or philosophy. Fewer than 10 percent of the participants got it right when Peter Cathcart Wason performed his 1966 study, the Wason Selection Task. But according to an essay by Bruce Schneier referencing the work of evolutionary psychologists Leda Cosmides and John Tooby, that number improves, by 65 to 80 percent "...when the rule has to do with cheating and privilege."
posted by fartknocker (35 comments total)

It's fascinating that people seem hard-wired to detect cheating and privilege, and that it can sort of be tested against other logical thought problems with such a simple evaluation.

It says a lot about how people are distrusting of our public policies and election processes and business culture nearly instinctually at this point. We're able to intuitively smell a rat in the system. And even if there's no actual cheating going on, the appearance of it erodes confidence in a system just as much as actual impropriety.
posted by hippybear at 10:02 AM on April 10, 2011

The conclusion that it's because it's about cheating isn't really supported by the evidence. My first hypothesis was "it's because they were emotionally involved in the outcome." Cheating is an emotionally charged situation. The first comment on the first link makes a similar objection:

The first test seems senseless or arbitrary. The second test does not. Resolving the skeleton of a scenario, implied by the isolated constraints, into a satisfactory, concrete situation is what is failing in the first test.
I'm sure plenty of subjects wondered "Why is car travel restricted into Boston? For how long?", and I'm sure they're misinterpreting the "if, then" as an equivalence. If it were a real situation, much more information would be required to make a decision about the fact that all Boston travel is by plane.
The reason they're thinking that is that they aren't abstracting the situation; they're looking for causes, effects, and correlations that seem familiar to them. I disagree with characterizing this as "cheating" in a "social exchange", based on the examples given. It's a simple matter of concreteness
posted by Obscure Reference at 10:12 AM on April 10, 2011

Here's an online interactive version of the test. Two are abstract and the final one might be construed as cheating.

Having read the essay and aced logic courses, I found it pretty straight forward, but this particular logic is very counter intuitive without a lot of practice. Or, apparently, the right scenario.
posted by pwnguin at 10:14 AM on April 10, 2011

Notice how quickly we were able to recognize that the evo-psych conclusion was arrived at through improper logic. QED!
posted by Sticherbeast at 10:14 AM on April 10, 2011 [9 favorites]

Interesting. I never had trouble grasping the confirmation bias with this test when I was a psych student -- it's as logical and straightforward as it gets.

Personalizing, and perhaps to a degree, infantilizing it into some sort of petty "that's not fair!" scenario seems to get people to take more notice whether someone is getting something they shouldn't be, but then you see all sorts of real life examples where people get away with the same sort of things the cheating experiments examined, and people don't see it.

The same underlying logic may be there, but for a lot of people, it's inconsistently applied, or it only goes so far and for certain situations...
posted by Alexandra Kitty at 10:16 AM on April 10, 2011

I have to imagine that 10% is in part due to the convoluted nature of the test. Adding the possibility of cheating into the equation means you're much more likely to look for loopholes, which brings to light the "trick question" nature of the test.
posted by o0o0o at 10:16 AM on April 10, 2011 [1 favorite]

I find this fascinating.

Am I right in thinking that the way to "cheat" is to redefine the cost for one group (or yourself) so that it is not in violation of the rule? I'm thinking about the recent financial crisis and the way that banks (and banking executives) seem to be held to a standard that is different to the hoi polloi. Or politicians ethical lapses compared to others in the community.

I wonder, if we are so tuned into cheating, how is cheating so common? Is it not cheating if you've successfully defined the rule as not being in violation, even if it "cheats" other people in the process? "Political bribes and kickbacks? Oh, that's just they way it is".
posted by qwip at 10:18 AM on April 10, 2011

I'd like to see this test repeated with the syllogism being "If she really loved you, she would cry when she heard you were dying." I'll bet respondents do as well in this version as in the cheating one.
posted by Obscure Reference at 10:24 AM on April 10, 2011 [7 favorites]

I think it's hard to separate the cheating aspect from the presentation of the problem. At the heart of this task, it's a question of applying a set of rules. And in most rule systems that people are familiar with, the people who break the rules are cheaters. So cheating is sort of an intrinsic part of the problem once you move past an abstract rule system to something relateable. If you kept the problem abstract and/or unfamiliar and added a cheating aspect, I think people would still have a hard time using their normal common-sense notions of rules to help them solve the problem. The key is that things like "eat vegetables before desert" and "only drink alcohol if you are over 21" are rules that we already know and can apply to these sorts of tests easily.
posted by burnmp3s at 10:24 AM on April 10, 2011

I've heard of a version along the lines of:

You're an alcohol control officer inspecting a bar, and you see (1) an adolescent and (2) a senior citizen (but you can't tell what either are drinking), and you also see (3) a patron drinking a beer and (4) a patron drinking a (non-alcoholic) energy drink.

Who do you have to check on?

I suspect the proportion of correct answers would be higher than when the subject is tested with cards and some abstract scenario. But does the bar problem really involve "cheating" as the essay's author might argue? If so, than cheating is being very broadly defined here.
posted by Mapes at 10:27 AM on April 10, 2011 [1 favorite]

I feel like this is more an issue of real life application. The vernacular of logic tests are alienating and scary to most people who aren't academics. I feel like cheating (whatever that specifically means) is just a way in which this abstract logic problem can be applied to real world logic. I don't think it means we're all hardwired cheats and liars.
posted by jenniferteeter at 10:32 AM on April 10, 2011 [1 favorite]

I'd like to see this test repeated with the syllogism being "If she really loved you, she would cry when she heard you were dying." I'll bet respondents do as well in this version as in the cheating one.

That might make it harder-- it sounds like a litmus test for determining if she loved you. I.e., if she is crying than that proves she loves you, which is not necessarily true.
posted by justkevin at 10:36 AM on April 10, 2011

Dear people who are taking the idea that the brain is 'hard-wired' to detect cheating seriously:

It's not. There are a few things the brain is hard-wired for. Keeping you breathing, for instance. But most things your brain does for you, it has to learn that shit. An example: you know how natural it is for you to look around and see things? Seems like your brain was just made for that, right? It wasn't. Before you even can be consciously aware of your environment, the areas of your cortex receiving information from your eyes have to reorganize - neurons have to make the right connections, re-wire themselves. Sure, there're some evolutionary bits that happened that govern how the eyes themselves get wired up to the brain, but anything after that point is mostly learned.

So the bottom line is, whenever someone tells you that the brain is hard-wired to do something, especially if it's something abstract like 'cheating', it is highly probable that person is full of shit.
posted by logicpunk at 10:38 AM on April 10, 2011 [1 favorite]

Alternatively, it could just mean that there are some linguistic contexts in which people tend to interpret "if" as a material conditional, and other linguistic contexts in which its most common vernacular meaning is different (from the wikipedia link).

That's it right there.
posted by yesster at 10:39 AM on April 10, 2011 [3 favorites]

Here's an online interactive version of the test. Two are abstract and the final one might be construed as cheating.

Damn I failed that! Got the first two wrong. The reason I got the third right wasn't cause it was to do with cheating but because it was an application of a real world problem (do I need to check what the underage person is drinking to know if they're drinking alcohol? YES)
posted by the mad poster! at 10:40 AM on April 10, 2011

"How the fuck you able to keep the count right you not able to do the book problem then?"
"Count be wrong, they fuck you up."

-The Wire, season 1, episode 8 "The Heavy Lifting"
posted by ceribus peribus at 10:43 AM on April 10, 2011 [13 favorites]

I think burnmp3s makes a good point. The problem is inherently about figuring out whether or not something has broken a rule, and that makes it difficult to come up with a scenario that is plausible but not in any way related to cheating or breaking social codes.

Maybe something like:

Imagine this: Our understanding is that a striped mother cat will always give birth to striped kittens. But we're not sure about this. We're trying to figure out if our assumption is correct. So here we have a striped mother cat, but we'd have to open this closed door if we want to see what its kittens look like. And here we have a brown-coated mother cat, and its kittens are behind this other closed door. And here, third, we have a litter of striped cats, and their mom is behind that third door. And lastly we have this litter of brown-coated cats, and their mom is behind that fourth door. Now, opening doors is hard work! Who wants to do more work than they need to! Think it through, and tell me which doors you'd need to open if you wanted to make sure our initial assumption was correct.

posted by nobody at 10:44 AM on April 10, 2011 [1 favorite]

So the bottom line is, whenever someone tells you that the brain is hard-wired to do something, especially if it's something abstract like 'cheating', it is highly probable that person is full of shit.
posted by logicpunk at 10:38 AM on April 10 [+] [!]


Wait wait wait I know . . . eponysterical! :D
posted by PhoBWanKenobi at 10:47 AM on April 10, 2011

Interesting. It makes perfect sense to me to say that logic evolved to detect cheaters. After all, when someone--say, three million years ago--says to you "There's something good to eat over the hill near the lions" it certainly would pay to have some way of telling whether he/she was telling the truth *other than visiting the lions.* Logic gives you that, at least in part.
posted by MarshallPoe at 11:02 AM on April 10, 2011

MarshallPoe, isn't there a much simpler way to understand why people get conceptual logic wrong but can apply it properly within a contextual instance?
posted by the mad poster! at 11:05 AM on April 10, 2011

Is this sort of a double?
posted by Chuckles at 11:06 AM on April 10, 2011

This data wades into the subjunctive conditional and related phenomena.
posted by Brian B. at 11:09 AM on April 10, 2011

Also, it doesn't improve by 80 percent, it improves to 80 percent.
posted by Chuckles at 11:13 AM on April 10, 2011

Nobody, Kitties!
posted by dibblda at 11:17 AM on April 10, 2011

I think what cheating gives people is the right mental framework for the problem. You're trying to test the violations of the rule, and for many easily constructed if-then scenarios, they're known (or assumed to be) true. With cheating, it's easier for people to question the truth of the implication as a whole, rather than applying the implication as truth to the problem.
posted by pwnguin at 11:36 AM on April 10, 2011

You're an alcohol control officer inspecting a bar, and you see (1) an adolescent and (2) a senior citizen (but you can't tell what either are drinking), and you also see (3) a patron drinking a beer and (4) a patron drinking a (non-alcoholic) energy drink.

Malcolm Gladwell talks about the Watson Selection Task in The Tipping Point, and uses this as the real life analogue. Unfortunately, it looks like the relevant passage isn't available on Google Books.
posted by alphanerd at 12:11 PM on April 10, 2011

OP probably should have linked to some primary sources. See this paper by Cosmides and Tooby, in which they test 10 (ten) alternative hypotheses:

Alternative (By-product) Hypotheses Eliminated

B1. That familiarity can explain the social contract effect.
B2. That social contract content merely activates the rules of inference of the propositional calculus (logic).
B3. That any problem involving payoffs will elicit the detection of logical violations.
B4. That permission schema theory can explain the social contract effect.
B5. That social contract content merely promotes “clear thinking.”
B6. That a content-independent deontic logic can explain social contract reasoning.
B7. That a single mechanism operates on all deontic rules involving subjective utilities.
B8. That relevance theory can explain social contract effects (see also Fiddick et al., 2000).
B9. That rational choice theory can explain social contract effects.
B10. That statistical learning produces the mechanisms that cause social contract
reasoning.

This bit is particularly relevant to some of the objections brought up in this thread:

The evidence supports social contract theory: Cheater detection occurs even when the social contract is wildly unfamiliar (Figure 20.3a). For example, the rule, “If a man eats cassava root, then he must have a tattoo on his face,” can be made to fit the social contract template by explaining that the people involved consider eating cassava root to be a benefit (the rule then implies that having a tattoo is the requirement an individual must satisfy to be eligible for that benefit). When given this context, this outlandish, culturally alien rule elicits the same high level of cheater detection as highly familiar social exchange rules. This surprising result has been replicated for many different unfamiliar rules (Cosmides, 1985,
1989; Cosmides & Tooby, 1992; Gigerenzer & Hug, 1992; Platt & Griggs, 1993).

posted by AceRock at 2:29 PM on April 10, 2011 [3 favorites]

So the bottom line is, whenever someone tells you that the brain is hard-wired to do something, especially if it's something abstract like 'cheating', it is highly probable that person is full of shit.

If your brain didn't have the capacity to do something, no amount of learning will allow your brain to do it. So yes, we do have to learn to see, but if we didn't have a capacity for vision already present in the brain at birth, no amount of training would allow it to learn to use the eyes. In the same way, if we didn't have the capacity to detect deception, no amount of study would allow us to learn about cheating.

I know what you're saying, but don't dismiss me as being full of shit simply because I didn't frame my statement in exactly the bio-developmental terms you require. Let's talk like people rather than being so dismissive.
posted by hippybear at 2:32 PM on April 10, 2011 [4 favorites]

"It's illegal to go to Boston, except by plane." Let's see how well we do spotting the cheaters now.
posted by Obscure Reference at 2:54 PM on April 10, 2011

"It's illegal to go to Boston, except by plane."

The answer involves "hoax devices," right?
posted by evidenceofabsence at 3:09 PM on April 10, 2011

If your brain didn't have the capacity to do something, no amount of learning will allow your brain to do it.

That's very different from saying something is hardwired, however.
posted by anigbrowl at 3:40 PM on April 10, 2011

Hippybear, I'm with you on the rudeness, but your rephrasing sort of nullifies the stronger sentiment you were expressing in your first comment. I mean, by the "if your brain didn't have the capacity to do something..." test, you could have been saying "It's fascinating that people seem hard-wired to understand relativity and quantum dynamics."

That said, the article AceRock linked to -- which I've only barely skimmed -- does seem much more convincing than the OP. This part caught my eye:

As predicted (D3), the mind’s automatically deployed definition of cheating is tied to the perspective you are taking (Gigerenzer & Hug, 1992). For example, consider the following social contract:

[1] If an employee is to get a pension, then that employee must have worked for the firm for over 10 years.

This rule elicits different answers depending on whether subjects are cued into the role of employer or employee. Those in the employer role look for cheating by employees, investigating cases of P and not-Q (employees with pensions; employees who have worked for fewer than 10 years). Those in the employee role look for cheating by employers, investigating cases of not-P and Q (employees with no pension; employees who have worked more than 10 years). Not-P & Q is correct if the goal is to find out whether the employer is cheating employees. But it is not
logically correct.

posted by nobody at 3:44 PM on April 10, 2011

nobody: Sure, I'll accept that. I still think that a lot of the things which are learned are only possible because the capacity for them is innate, and maybe the word "hardwired" was a bit too extreme. But I do also think that there are a lot of very universal traits within human consciousness which cannot be easily accounted for via Skinner-box experiments. And amongst those lie things like the ability to tell truth from fiction and cheaters from those playing by the rules. We may lack the ability to really quantify exactly why these traits are universal, but they seem to exist across all cultures and across human history, so there has to be something more going on than simply "this is a learned thing".

So, I'll accept that my first statement may have been a bit too strongly worded, but I won't back down from postulating that concepts such as justice and fair play and cheating are part of the human psychological makeup, and that the way the human brain is formed plays a large if not dominant role in determining exactly what human psychological makeup is.
posted by hippybear at 3:52 PM on April 10, 2011

hippybear: sorry if it were ill-phrased; I wasn't specifically saying you were full of shit - that comment was directed specifically at the author of the linked post, and I stand by it. It's a real problem, separating out evolutionary influences on cognition from cultural influences, and people making statements about how brains are hard-wired to do certain things (presumably by some putative evolutionary mechanism) is misleading, and can be used to justify a lot of shitty things (e.g., males are genetically predisposed to be better at abstract reasoning than females).
posted by logicpunk at 5:15 PM on April 10, 2011

This could all be so simple.

Given any statement of truth the only statement you can make that is 100% correct—always, verifiably, irrefutably, and without question—is the contrapositive.

statement: if A, then B
contrapositive: if not B, then not A

See how that works? You switch around the if… then and you add a not in front of each part.

This is the only thing you have to learn with logic. If people were just taught this single fact—hell, spend a whole year on it!—but this single fact would eliminate 99% of the pain and suffering in the world. 99% of the strife. 99 kittens out of 100 would be less euthanized. It's a tragic tragedy of epic proportion, really.
posted by Civil_Disobedient at 5:56 AM on April 11, 2011